We have been without news of OpenGL 3.0 for nearly 9 months now, a specification was promised for September 2007 at Siggraph last year and then no news after an ARB member reported that the spec was not ready since there was some unresolved issues they had to address. This situation started to becomes really worrying and a lot of speculations have been made on this delay (see opengl.org forum). The most likely thing is that there have been some disagreements inside ARB members that delayed the spec release.

But recently we had two comforting news tending to prove that OpenGL 3.0 is not dead. The first thing is the creation of the OpenGL 3.0 website with announcements for the Siggraph OpenGL BOF :

    - "OpenGL 3.0 Specification Overview"

    - "OpenGL hardware and driver plans - AMD, Intel, NVIDIA"

    - "Developer's perspective on OpenGL 3.0 "

The second think is a leeked picture of an NVidia presentation slide about their future drivers release (codename Big Bang 2, see the news here) where can distinguiche a reference to OpenGL 3.0 support for September. This tend to prove that the specification is now almost done and the IHV are already working on implementation.

I will be at Siggraph this year and I will be at the OpenGL BOF, I hope I will be able to the the final spec here ! 

There is a very interesting article on G80/G92 hardware architecture just published by NVIDIA guys (hardware architects) in the March/April edition of IEEE Micro:

    Erik Lindholm, NVIDIA
    John Nickolls, NVIDIA
    Stuart Oberman, NVIDIA
    John Montrym, NVIDIA

The full article can be downloaded here: http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4523358

Most of the informations can be found or deduced from the CUDA manual but there is also few details about hardware model in this article that was not given before as directly by NVIDIA. Especially, there is mention of Fragments grouped into 2x2 to be scheduled into Warps, a thing we deduced with our G80 hardware analysis (http://www.icare3d.org/GPU/CN08).There is also the firstpicture of the G80 die publicly avalable (hum ok, not very clear and in grayscale) !

The the French national High-Performance Computing organization (GENCI) composed among others by the French Atomic Energy Authority (CEA) just order to Bull one of the first supercomputer (a large-scale PC cluster) using an hybrid architecture exploiting GPUs as high performance computational resource. This cluster will be part of CCRT (the Center for Research and Technology computing) and installed in the the French Atomic Energy Authority's Directorate of Military Affairs center (CEA/DAM) of Bruyères le Châtel.

This cluster will use 48 NVidia Tesla GPU modules that are very likely to be based on the next NVIDIA processor generation, the GT200, especially because this generation will support double precision (64 bits floating points) in hardware (but the question is still the cost of this). Each GPU module is said to be composed of 512 cores that let supposed that they will be composed of two GT200 GPUs providing 256 Stream Processors each. Since I already heard rumors about a 240 Stream Processors configuration for the GT200, it is likely to be composed of 16 Multi Processors of 16 Stream Processors each. This would allow a GTX version composed of 15 MP and an ultra version composed of 16MP.

I know quite well the CEA/DAM center since I did an internship here a few years ago, and I know a few guys here who I am sure will be very happy with this new very cool toy ! ;-D

Sources:

- http://www.pcinpact.com/actu/news/43165-premier-supercalculateur-GPU-France-Tesla.htm
- http://www.wcm.bull.com/internet/pr/rend.jsp?DocId=350335&lang=fr

Nvidia has released a few days ago an online version of  GPU Gems: Programming Techniques, Tips, and Tricks for Real-Time Graphics, the first tome of what I consider as one of the best GPU book series.

GPU Gems is a compilation of articles covering practical real-time graphics techniques arising from the research and practice of cutting-edge developers. It focuses on the programmable graphics pipeline available in today's graphics processing units (GPUs) and highlights quick and dirty tricks used by leading developers, as well as fundamental, performance-conscious techniques for creating advanced visual effects. The contributors and editors, collectively, bring countless years of experience to enlighten and propel the reader into the fascinating world of programmable real-time graphics.

The book can be read here: http://developer.nvidia.com/gpugems.

    For today GPU generation, we know a lot more about the hardware mechanism than for previous GPU generation, in particular thanks to the window open on this hardware by Cuda. But many hardware details are still hidden to the programmer, in particular mechanisms used for primitives rasterisation and fragments shading. As understanding how fragments are scheduled among the G80 processing units is a critical points for the research we do with Fabrice Neyret as part of my PhD, I wrote a small program allowing to investigate this point. 

     We also wrote a document that presents our motivations for this work, the experiments we made yet and the results and answers we get: http://www.icare3d.org/GPU/CN08

    We hope this document will give you useful informations. We don't want it to be closed and we want to make it evolve through new experiments and also thanks to exterior comments you can let at the bottom of this page.

 ---=== Understanding G80 behavior and performances ===---
Cyril Crassin, Fabrice Neyret LJK-INRIA, Grenoble, France
This e-mail address is being protected from spam bots, you need JavaScript enabled to view it This e-mail address is being protected from spam bots, you need JavaScript enabled to view it

http://www-evasion.imag.fr/GPU/CN08

This is an evolving document (corrections, additions): please link it, but don't copy it !

This is also an open document and we hope to engage discussions on the forum here.

version: March, 8 2008

    But till now, very few examples and tutorials on theses functionalities are available. The only tutorial I have been able to find is the one from Xie Yongming, it's a good start but unfortunately its source code is unavailable.

    This simple example will show how to use GLSL Geometry Shaders (EXT_geometry_shader4 and EXT_gpu_shader4) and the new integer texture formats (EXT_texture_integer) extensions to implement the Marching Cubes algorithm entirely on the GPU. The implementation I will present is probably not the most efficient but I think it is a good example, showing the usage of the new shader stage, new texture formats and binary operations within shader.

Alors que NVidia ne fournit pas encore de pilote pour DirectX 10 (donc windows Vista, seul système à supporter cette version de DirectX qui introduit l’ensemble des nouvelles fonctionnalités), il propose déjà une série d’extensions OpenGL exposant toutes les nouvelles capacités du GPU.

Je vais décrire rapidement ce qui me semble intéressant dans ce que j’ai lu de ces extensions dans les specs de NVidia (http://developer.nvidia.com/ object/nvidia_opengl_specs.html ).

Mercredi a été lancé le G80 ou GeForce 8800, nouveau GPU de chez NVidia. Il s’agit du premier GPU supportant les nouvelles fonctionnalités de DirectX 10 Il est doté pour cela d’une architecture totalement nouvelle sur laquelle très peu d’informations n’avaient filtré jusqu’a sa sortie officielle.

Je vais faire une petite review de ce GPU en commençant par décrire rapidement l’architecture du GPU. Ensuite je présenterai ce qui me semble intéressant et ce que j’ai pour l’instant découvert dans les nouvelles extensions OpenGL que NVidia propose avec.